Data processing method and device, equipment, storage medium and program product

Abstract

The invention provides a data processing method and device, equipment, a storage medium and a program product. The method comprises the steps of obtaining first index data corresponding to a monitoring index; a plurality of data processing algorithms are determined, the algorithm complexity of the data processing algorithms is smaller than or equal to preset complexity, and the data processing algorithms are used for performing data decomposition processing on data; and determining first historical index data corresponding to the first index data, and processing the first index data and the first historical index data through the plurality of data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring index. And the data processing efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of computers, and in particular to a data processing method, apparatus, device, storage medium, and program product. Background Art

[0002] Performing seasonal-trend decomposition (STD) on periodic time series data can yield trend data, periodic data, and residual data corresponding to the periodic time series data. Trend data, periodic data, and residual data facilitate further analysis of periodic time series data.

[0003] Currently, STD processing of periodic time series data can be performed using robust seasonal and trend decomposition using loess (RobustSTL) or single-shot time series (OneShotSTL) algorithms. However, these algorithms all include an optimization process based on the L1 norm, which is highly complex and results in low data processing efficiency. Summary of the Invention

[0004] Various aspects of the present application provide a data processing method, apparatus, device, storage medium, and program product to improve data processing efficiency.

[0005] In a first aspect, an embodiment of the present application provides a data processing method, comprising:

[0006] Obtaining first indicator data corresponding to the monitoring indicator;

[0007] Determining a plurality of data processing algorithms, wherein the algorithm complexity of the data processing algorithms is less than or equal to a preset complexity, and the plurality of data processing algorithms are used to perform data decomposition processing on the data;

[0008] Determine first historical indicator data corresponding to the first indicator data, and process the first indicator data and the first historical indicator data through the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator.

[0009] In one possible implementation, processing the first indicator data and the first historical indicator data by the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator includes:

[0010] Determining a first data volume of historical indicator data corresponding to the monitoring indicator;

[0011] If the first data volume is less than or equal to M, then when M indicator data corresponding to the monitoring indicator are obtained, the M indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data, and the M indicator data include the first indicator data and the first historical indicator data;

[0012] If the first data volume is greater than M, the first indicator data and M-1 first historical indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data.

[0013] In one possible implementation, processing the M indicator data using the multiple data processing algorithms to obtain the trend data and / or the periodic data includes:

[0014] Determining at least one second indicator data from the M indicator data, where the second indicator data is indicator data that undergoes a jump among the M indicator data;

[0015] According to the at least one second indicator data, the M indicator data are segmented to obtain a plurality of indicator data sequences;

[0016] Determine the trend value of each indicator data in each indicator data series, and determine the period value of each indicator data by using a non-local periodic filtering algorithm;

[0017] Determine that the trend data includes the trend value of each indicator data in the M indicator data, and determine that the periodic data includes the period value of each indicator data in the M indicator data; wherein the multiple data processing algorithms include the algorithms used in the above steps.

[0018] In a possible implementation, determining at least one second indicator data from the M indicator data includes:

[0019] Determining a plurality of third indicator data from the M indicator data, wherein the third indicator data is data in a middle portion of the M indicator data;

[0020] For any third indicator data, determining a jump value of the third indicator data, where the jump value is used to indicate a difference between future indicator data and historical indicator data of the third indicator data, where a generation time of the future indicator data is after a generation time of the third indicator data;

[0021] Determine, according to the jump amount of the third indicator data, candidate indicator data from the plurality of third indicator data, wherein the jump amount of the candidate indicator data is a local extreme value point in the jump amount of the plurality of third indicator data;

[0022] For any one candidate indicator data, determine the jump value threshold corresponding to the candidate indicator data according to the M indicator data;

[0023] Indicator data whose absolute value of the jump amount is greater than the corresponding jump amount threshold in the candidate indicator data is determined as the at least one second indicator data.

[0024] In a possible implementation, for any indicator data sequence, determining a trend value of each indicator data in the indicator data sequence includes:

[0025] If the number of indicator data included in the indicator data sequence is less than or equal to a first threshold value T, then determining the average value of each indicator data in the indicator data sequence as the trend value of each indicator data in the indicator data sequence;

[0026] If the number of indicator data included in the indicator data sequence is greater than the first threshold, determining a first subsequence and a second subsequence in the indicator data sequence, and determining a trend value of each indicator data in the first subsequence based on each indicator data in the indicator data sequence, and determining the trend value of each indicator data in the second subsequence as an average value of each indicator data in the first subsequence;

[0027] The second subsequence includes the last T-1 indicator data in the indicator data sequence, the first subsequence includes other indicator data in the indicator data sequence except the second subsequence, and T is an integer greater than 1.

[0028] In a possible implementation, for any indicator data in any indicator data sequence, determining a period value of the indicator data by using a non-local periodic filtering algorithm includes:

[0029] Determining, among the M indicator data, a plurality of first reference indicator data corresponding to the indicator data;

[0030] Determining a weight value of each first reference indicator data;

[0031] The period value of the indicator data is determined according to the weight value of each first reference indicator data, each first reference indicator data, and the trend value of each first reference indicator data.

[0032] In one possible implementation, processing the first indicator data and M-1 first historical indicator data by the multiple data processing algorithms to obtain the trend data and / or the periodic data includes:

[0033] Determining a predicted value corresponding to the first indicator data, and judging whether the first indicator data is outlier data based on the first indicator data and the predicted value;

[0034] If yes, determining a first number of continuous outlier data, and determining the trend data and / or the periodic data based on the first number, the first indicator data, and M-1 first historical indicator data;

[0035] If not, the average value of the first indicator data and the M-1 first historical indicator data is determined as the trend value of the first indicator data, and the period value of the first indicator data is determined by the non-local periodic filtering algorithm, wherein the trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0036] In a possible implementation, determining the trend data and / or the periodic data according to the first quantity, the first indicator data, and M-1 first historical indicator data includes:

[0037] Determining whether the first number is greater than or equal to a preset L, where L is an integer greater than or equal to 1;

[0038] If so, then determine the L-1 second historical indicator data before the first indicator data among the M-1 first historical indicator data, and determine a target trend value based on the first indicator data and the L-1 second historical indicator data, and update the trend values of the first indicator data and the L-1 second historical indicator data to the target trend value, and determine the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0039] If not, determining the average value of the predicted value and the M-1 first historical indicator data as the trend value of the first indicator data, and determining the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0040] The trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0041] In a possible implementation, after determining the target trend value based on the first indicator data and the L-1 second historical indicator data, the method further includes:

[0042] Updating the first indicator data and the L-1 second historical indicator data to the target trend value; and

[0043] For any other historical indicator data among the M-1 historical indicator data except the L-1 second historical indicator data, the update increment is determined based on the target trend data and the trend value of the Lth historical indicator data before the first indicator data, and the historical indicator data is updated to the sum of the historical indicator data and the update increment.

[0044] In one possible implementation, determining the predicted value corresponding to the first indicator data includes:

[0045] Determining, from the plurality of historical indicator data of the indicator data, a plurality of second reference indicator data corresponding to the first indicator data and a first trend value of indicator data preceding the first indicator data;

[0046] The predicted value is determined according to the first indicator data, the period value of each second reference indicator data, and the first trend value.

[0047] In a possible implementation, the trend data includes trend values of a plurality of historical indicator data corresponding to the monitoring indicator, and the periodic data includes period values of the plurality of historical indicator data; the method further includes:

[0048] For any historical indicator data, the sum of the trend value and the period value of the historical indicator data is determined as the to-be-processed value corresponding to the historical indicator data;

[0049] Determine the difference between the historical indicator data and the value to be processed as the residual value corresponding to the historical indicator data;

[0050] Determine that the residual data corresponding to the monitoring indicator includes the residual value corresponding to each historical indicator data.

[0051] In a second aspect, an embodiment of the present application provides a data processing method, including:

[0052] Obtaining first indicator data corresponding to the monitoring indicator, where the first indicator data is data generated by the operation of the device;

[0053] Determining a plurality of data processing algorithms, wherein the algorithm complexity of the data processing algorithms is less than or equal to a preset complexity, and the plurality of data processing algorithms are used to perform data decomposition processing on the data;

[0054] Determine first historical indicator data corresponding to the first indicator data, and perform data decomposition processing on the first indicator data and the first historical indicator data through the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator.

[0055] In one possible implementation, processing the first indicator data and the first historical indicator data by the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator includes:

[0056] Determining a first data volume of historical indicator data corresponding to the monitoring indicator;

[0057] If the first data volume is less than or equal to M, then when M indicator data corresponding to the monitoring indicator are obtained, the M indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data, and the M indicator data include the first indicator data and the first historical indicator data;

[0058] If the first data volume is greater than M, the first indicator data and M-1 first historical indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data.

[0059] In a possible implementation, processing the M indicator data using the multiple data processing algorithms to obtain the trend data and / or the periodic data includes:

[0060] Determining at least one second indicator data from the M indicator data, where the second indicator data is indicator data that undergoes a jump among the M indicator data;

[0061] According to the at least one second indicator data, the M indicator data are segmented to obtain a plurality of indicator data sequences;

[0062] Determine the trend value of each indicator data in each indicator data series, and determine the period value of each indicator data by using a non-local periodic filtering algorithm;

[0063] Determine that the trend data includes the trend value of each indicator data in the M indicator data, and determine that the periodic data includes the period value of each indicator data in the M indicator data; wherein the multiple data processing algorithms include the algorithms used in the above steps.

[0064] In a possible implementation, determining at least one second indicator data from the M indicator data includes:

[0065] Determining a plurality of third indicator data from the M indicator data, wherein the third indicator data is data in a middle portion of the M indicator data;

[0066] For any third indicator data, determining a jump value of the third indicator data, where the jump value is used to indicate a difference between future indicator data and historical indicator data of the third indicator data, where a generation time of the future indicator data is after a generation time of the third indicator data;

[0067] Determine, according to the jump amount of the third indicator data, candidate indicator data from the plurality of third indicator data, wherein the jump amount of the candidate indicator data is a local extreme value point in the jump amount of the plurality of third indicator data;

[0068] For any one candidate indicator data, determine the jump value threshold corresponding to the candidate indicator data according to the M indicator data;

[0069] Indicator data whose absolute value of the jump amount is greater than the corresponding jump amount threshold in the candidate indicator data is determined as the at least one second indicator data.

[0070] In a possible implementation, for any indicator data sequence, determining a trend value of each indicator data in the indicator data sequence includes:

[0071] If the number of indicator data included in the indicator data sequence is less than or equal to a first threshold value T, then determining the average value of each indicator data in the indicator data sequence as the trend value of each indicator data in the indicator data sequence;

[0072] If the number of indicator data included in the indicator data sequence is greater than the first threshold, determining a first subsequence and a second subsequence in the indicator data sequence, and determining a trend value of each indicator data in the first subsequence based on each indicator data in the indicator data sequence, and determining the trend value of each indicator data in the second subsequence as an average value of each indicator data in the second subsequence;

[0073] The second subsequence includes the last T-1 indicator data in the indicator data sequence, the first subsequence includes other indicator data in the indicator data sequence except the second subsequence, and T is an integer greater than 1.

[0074] In a possible implementation, for any indicator data in any indicator data sequence, determining a period value of the indicator data by using a non-local periodic filtering algorithm includes:

[0075] Determining, among the M indicator data, a plurality of first reference indicator data corresponding to the indicator data;

[0076] Determining a weight value of each first reference indicator data;

[0077] The period value of the indicator data is determined according to the weight value of each first reference indicator data, each first reference indicator data, and the trend value of each first reference indicator data.

[0078] In one possible implementation, processing the first indicator data and M-1 first historical indicator data by the multiple data processing algorithms to obtain the trend data and / or the periodic data includes:

[0079] Determining a predicted value corresponding to the first indicator data, and judging whether the first indicator data is outlier data based on the first indicator data and the predicted value;

[0080] If yes, determining a first number of continuous outlier data, and determining the trend data and / or the periodic data based on the first number, the first indicator data, and M-1 first historical indicator data;

[0081] If not, the average value of the first indicator data and the M-1 first historical indicator data is determined as the trend value of the first indicator data, and the period value of the first indicator data is determined by the non-local periodic filtering algorithm, wherein the trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0082] In a possible implementation, determining the trend data and / or the periodic data according to the first quantity, the first indicator data, and M-1 first historical indicator data includes:

[0083] Determining whether the first number is greater than or equal to a preset L, where L is an integer greater than or equal to 1;

[0084] If so, then determine the L-1 second historical indicator data before the first indicator data among the M-1 first historical indicator data, and determine a target trend value based on the first indicator data and the L-1 second historical indicator data, and update the trend values of the first indicator data and the L-1 second historical indicator data to the target trend value, and determine the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0085] If not, determining the average value of the predicted value and the M-1 first historical indicator data as the trend value of the first indicator data, and determining the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0086] The trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0087] In a possible implementation, after determining the target trend value based on the first indicator data and the L-1 second historical indicator data, the method further includes:

[0088] Updating the first indicator data and the L-1 second historical indicator data to the target trend value; and

[0089] For any other historical indicator data among the M-1 historical indicator data except the L-1 second historical indicator data, the update increment is determined based on the target trend data and the trend value of the Lth historical indicator data before the first indicator data, and the historical indicator data is updated to the sum of the historical indicator data and the update increment.

[0090] In one possible implementation, determining the predicted value corresponding to the first indicator data includes:

[0091] Determining, from the plurality of historical indicator data of the indicator data, a plurality of second reference indicator data corresponding to the first indicator data and a first trend value of indicator data preceding the first indicator data;

[0092] The predicted value is determined according to the first indicator data, the period value of each second reference indicator data, and the first trend value.

[0093] In a possible implementation, the trend data includes trend values of a plurality of historical indicator data corresponding to the monitoring indicator, and the periodic data includes period values of the plurality of historical indicator data; the method further includes:

[0094] For any historical indicator data, the sum of the trend value and the period value of the historical indicator data is determined as the to-be-processed value corresponding to the historical indicator data;

[0095] Determine the difference between the historical indicator data and the value to be processed as the residual value corresponding to the historical indicator data;

[0096] Determine that the residual data corresponding to the monitoring indicator includes the residual value corresponding to each historical indicator data.

[0097] In a third aspect, an embodiment of the present application provides a data processing device, comprising: an acquisition module, a determination module, and a processing module, wherein:

[0098] The acquisition module is used to obtain first indicator data corresponding to the monitoring indicator;

[0099] The determination module is used to determine multiple data processing algorithms, the algorithm complexity of the data processing algorithms is less than or equal to the preset complexity, and the multiple data processing algorithms are used to perform data decomposition processing on the data;

[0100] The determining module is further configured to determine first historical indicator data corresponding to the first indicator data;

[0101] The processing module is used to process the first indicator data and the first historical indicator data through the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator.

[0102] In a possible implementation, the processing module is specifically configured to:

[0103] Determining a first data volume of historical indicator data corresponding to the monitoring indicator;

[0104] If the first data volume is less than or equal to M, then when M indicator data corresponding to the monitoring indicator are obtained, the M indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data, and the M indicator data include the first indicator data and the first historical indicator data;

[0105] If the first data volume is greater than M, the first indicator data and M-1 first historical indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data.

[0106] In a possible implementation, the processing module is specifically configured to:

[0107] Determining at least one second indicator data from the M indicator data, where the second indicator data is indicator data that undergoes a jump among the M indicator data;

[0108] According to the at least one second indicator data, the M indicator data are segmented to obtain a plurality of indicator data sequences;

[0109] Determine the trend value of each indicator data in each indicator data series, and determine the period value of each indicator data by using a non-local periodic filtering algorithm;

[0110] Determine that the trend data includes the trend value of each indicator data in the M indicator data, and determine that the periodic data includes the period value of each indicator data in the M indicator data; wherein the multiple data processing algorithms include the algorithms used in the above steps.

[0111] In a possible implementation, the processing module is specifically configured to:

[0112] Determining a plurality of third indicator data from the M indicator data, wherein the third indicator data is data in a middle portion of the M indicator data;

[0113] For any third indicator data, determining a jump value of the third indicator data, where the jump value is used to indicate a difference between future indicator data and historical indicator data of the third indicator data, where a generation time of the future indicator data is after a generation time of the third indicator data;

[0114] Determine, according to the jump amount of the third indicator data, candidate indicator data from the plurality of third indicator data, wherein the jump amount of the candidate indicator data is a local extreme value point in the jump amount of the plurality of third indicator data;

[0115] For any one candidate indicator data, determine the jump value threshold corresponding to the candidate indicator data according to the M indicator data;

[0116] Indicator data whose absolute value of the jump amount is greater than the corresponding jump amount threshold in the candidate indicator data is determined as the at least one second indicator data.

[0117] In a possible implementation, for any indicator data sequence, the processing module is specifically configured to:

[0118] If the number of indicator data included in the indicator data sequence is less than or equal to a first threshold value T, then determining the average value of each indicator data in the indicator data sequence as the trend value of each indicator data in the indicator data sequence;

[0119] If the number of indicator data included in the indicator data sequence is greater than the first threshold, determining a first subsequence and a second subsequence in the indicator data sequence, and determining a trend value of each indicator data in the first subsequence based on each indicator data in the indicator data sequence, and determining the trend value of each indicator data in the second subsequence as an average value of each indicator data in the first subsequence;

[0120] The second subsequence includes the last T-1 indicator data in the indicator data sequence, the first subsequence includes other indicator data in the indicator data sequence except the second subsequence, and T is an integer greater than 1.

[0121] In a possible implementation, for any indicator data in any indicator data sequence, the processing module is specifically configured to:

[0122] Determining, among the M indicator data, a plurality of first reference indicator data corresponding to the indicator data;

[0123] Determining a weight value of each first reference indicator data;

[0124] The period value of the indicator data is determined according to the weight value of each first reference indicator data, each first reference indicator data, and the trend value of each first reference indicator data.

[0125] In a possible implementation, the processing module is specifically configured to:

[0126] Determining a predicted value corresponding to the first indicator data, and judging whether the first indicator data is outlier data based on the first indicator data and the predicted value;

[0127] If yes, determining a first number of continuous outlier data, and determining the trend data and / or the periodic data based on the first number, the first indicator data, and M-1 first historical indicator data;

[0128] If not, the average value of the first indicator data and the M-1 first historical indicator data is determined as the trend value of the first indicator data, and the period value of the first indicator data is determined by the non-local periodic filtering algorithm, wherein the trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0129] In a possible implementation, the processing module is specifically configured to:

[0130] Determining whether the first number is greater than or equal to a preset L, where L is an integer greater than or equal to 1;

[0131] If so, then determine the L-1 second historical indicator data before the first indicator data among the M-1 first historical indicator data, and determine a target trend value based on the first indicator data and the L-1 second historical indicator data, and update the trend values of the first indicator data and the L-1 second historical indicator data to the target trend value, and determine the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0132] If not, determining the average value of the predicted value and the M-1 first historical indicator data as the trend value of the first indicator data, and determining the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0133] The trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0134] In a possible implementation, the data processing device may further include an update module, wherein:

[0135] The updating module is used to update the first indicator data and the L-1 second historical indicator data to the target trend value;

[0136] For any other historical indicator data among the M-1 historical indicator data except the L-1 second historical indicator data, the determining module is further configured to determine an update increment based on the target trend data and a trend value of the Lth historical indicator data before the first indicator data;

[0137] The updating module is further configured to update the historical indicator data to the sum of the historical indicator data and the update increment.

[0138] In a possible implementation, the processing module is specifically configured to:

[0139] Determining, from the plurality of historical indicator data of the indicator data, a plurality of second reference indicator data corresponding to the first indicator data and a first trend value of indicator data preceding the first indicator data;

[0140] The predicted value is determined according to the first indicator data, the period value of each second reference indicator data, and the first trend value.

[0141] In a possible implementation, the trend data includes trend values of a plurality of historical indicator data corresponding to the monitoring indicator, and the periodic data includes period values of the plurality of historical indicator data; the determination module is further configured to:

[0142] For any historical indicator data, the sum of the trend value and the period value of the historical indicator data is determined as the to-be-processed value corresponding to the historical indicator data;

[0143] Determine the difference between the historical indicator data and the value to be processed as the residual value corresponding to the historical indicator data;

[0144] Determine that the residual data corresponding to the monitoring indicator includes the residual value corresponding to each historical indicator data.

[0145] In a fourth aspect, an embodiment of the present application provides a data processing device, comprising: an acquisition module, a determination module, and a processing module, wherein:

[0146] The acquisition module is used to acquire first indicator data corresponding to the monitoring indicator, where the first indicator data is data generated by the operation of the device;

[0147] The determination module is used to determine multiple data processing algorithms, the algorithm complexity of the data processing algorithms is less than or equal to the preset complexity, and the multiple data processing algorithms are used to perform data decomposition processing on the data;

[0148] The determining module is further configured to determine first historical indicator data corresponding to the first indicator data;

[0149] The processing module is used to process the first indicator data and the first historical indicator data through the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator.

[0150] In a possible implementation, the processing module is specifically configured to:

[0151] Determining a first data volume of historical indicator data corresponding to the monitoring indicator;

[0152] If the first data volume is less than or equal to M, then when M indicator data corresponding to the monitoring indicator are obtained, the M indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data, and the M indicator data include the first indicator data and the first historical indicator data;

[0153] If the first data volume is greater than M, the first indicator data and M-1 first historical indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data.

[0154] In a possible implementation, the processing module is specifically configured to:

[0155] Determining at least one second indicator data from the M indicator data, where the second indicator data is indicator data that undergoes a jump among the M indicator data;

[0156] According to the at least one second indicator data, the M indicator data are segmented to obtain a plurality of indicator data sequences;

[0157] Determine the trend value of each indicator data in each indicator data series, and determine the period value of each indicator data by using a non-local periodic filtering algorithm;

[0158] Determine that the trend data includes the trend value of each indicator data in the M indicator data, and determine that the periodic data includes the period value of each indicator data in the M indicator data; wherein the multiple data processing algorithms include the algorithms used in the above steps.

[0159] In a possible implementation, the processing module is specifically configured to:

[0160] Determining a plurality of third indicator data from the M indicator data, wherein the third indicator data is data in a middle portion of the M indicator data;

[0161] For any third indicator data, determining a jump value of the third indicator data, where the jump value is used to indicate a difference between future indicator data and historical indicator data of the third indicator data, where a generation time of the future indicator data is after a generation time of the third indicator data;

[0162] Determine, according to the jump amount of the third indicator data, candidate indicator data from the plurality of third indicator data, wherein the jump amount of the candidate indicator data is a local extreme value point in the jump amount of the plurality of third indicator data;

[0163] For any one candidate indicator data, determine the jump value threshold corresponding to the candidate indicator data according to the M indicator data;

[0164] Indicator data whose absolute value of the jump amount is greater than the corresponding jump amount threshold in the candidate indicator data is determined as the at least one second indicator data.

[0165] In a possible implementation, for any indicator data sequence, the processing module is specifically configured to:

[0166] If the number of indicator data included in the indicator data sequence is less than or equal to a first threshold value T, then determining the average value of each indicator data in the indicator data sequence as the trend value of each indicator data in the indicator data sequence;

[0167] If the number of indicator data included in the indicator data sequence is greater than the first threshold, determining a first subsequence and a second subsequence in the indicator data sequence, and determining a trend value of each indicator data in the first subsequence based on each indicator data in the indicator data sequence, and determining the trend value of each indicator data in the second subsequence as an average value of each indicator data in the first subsequence;

[0168] The second subsequence includes the last T-1 indicator data in the indicator data sequence, the first subsequence includes other indicator data in the indicator data sequence except the second subsequence, and T is an integer greater than 1.

[0169] In a possible implementation, for any indicator data in any indicator data sequence, the processing module is specifically configured to:

[0170] Determining, among the M indicator data, a plurality of first reference indicator data corresponding to the indicator data;

[0171] Determining a weight value of each first reference indicator data;

[0172] The period value of the indicator data is determined according to the weight value of each first reference indicator data, each first reference indicator data, and the trend value of each first reference indicator data.

[0173] In a possible implementation, the processing module is specifically configured to:

[0174] Determining a predicted value corresponding to the first indicator data, and judging whether the first indicator data is outlier data based on the first indicator data and the predicted value;

[0175] If yes, determining a first number of continuous outlier data, and determining the trend data and / or the periodic data based on the first number, the first indicator data, and M-1 first historical indicator data;

[0176] If not, the average value of the first indicator data and the M-1 first historical indicator data is determined as the trend value of the first indicator data, and the period value of the first indicator data is determined by the non-local periodic filtering algorithm, wherein the trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0177] In a possible implementation, the processing module is specifically configured to:

[0178] Determining whether the first number is greater than or equal to a preset L, where L is an integer greater than or equal to 1;

[0179] If so, then determine the L-1 second historical indicator data before the first indicator data among the M-1 first historical indicator data, and determine a target trend value based on the first indicator data and the L-1 second historical indicator data, and update the trend values of the first indicator data and the L-1 second historical indicator data to the target trend value, and determine the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0180] If not, determining the average value of the predicted value and the M-1 first historical indicator data as the trend value of the first indicator data, and determining the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0181] The trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0182] In a possible implementation, the data processing device may further include an update module, wherein:

[0183] The updating module is used to update the first indicator data and the L-1 second historical indicator data to the target trend value;

[0184] For any other historical indicator data among the M-1 historical indicator data except the L-1 second historical indicator data, the determining module is further configured to determine an update increment based on the target trend data and a trend value of the Lth historical indicator data before the first indicator data;

[0185] The updating module is further configured to update the historical indicator data to the sum of the historical indicator data and the update increment.

[0186] In a possible implementation, the processing module is specifically configured to:

[0187] Determining, from the plurality of historical indicator data of the indicator data, a plurality of second reference indicator data corresponding to the first indicator data and a first trend value of indicator data preceding the first indicator data;

[0188] The predicted value is determined according to the first indicator data, the period value of each second reference indicator data, and the first trend value.

[0189] In a possible implementation, the trend data includes trend values of a plurality of historical indicator data corresponding to the monitoring indicator, and the periodic data includes period values of the plurality of historical indicator data; the determination module is further configured to:

[0190] For any historical indicator data, the sum of the trend value and the period value of the historical indicator data is determined as the to-be-processed value corresponding to the historical indicator data;

[0191] Determine the difference between the historical indicator data and the value to be processed as the residual value corresponding to the historical indicator data;

[0192] Determine that the residual data corresponding to the monitoring indicator includes the residual value corresponding to each historical indicator data.

[0193] In a fifth aspect, an embodiment of the present application provides a data processing device, including: a memory and a processor;

[0194] The memory stores computer-executable instructions;

[0195] The processor executes the computer-executable instructions stored in the memory, so that the data processing performs the method according to any one of the first aspect or the second aspect.

[0196] In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method described in either the first aspect or the second aspect.

[0197] In a seventh aspect, an embodiment of the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements the method shown in either the first aspect or the second aspect.

[0198] The embodiments of the present application provide a data processing method, apparatus, device, storage medium, and program product, wherein the data processing device can obtain first indicator data corresponding to a monitoring indicator and determine multiple data processing algorithms; the first indicator data and first historical indicator data can be processed by multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator. In the above method, the complexity of the multiple data processing algorithms is low, so that the complexity of decomposing the indicator data (including the first indicator data and the first historical indicator data) into trend data and / or periodic data corresponding to the monitoring indicator is low, thereby improving data processing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0199] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0200] Figure 1 A schematic diagram of a scenario provided for an exemplary embodiment of the present application;

[0201] Figure 2 A flowchart of a data processing method provided by an exemplary embodiment of the present application;

[0202] Figure 3 A schematic diagram of the quantity of indicator data in a sliding window provided by an exemplary embodiment of the present application;

[0203] Figure 4 A flowchart of another data processing method provided for an exemplary embodiment of the present application;

[0204] Figure 5 A schematic diagram of an indicator data sequence provided for an exemplary embodiment of the present application;

[0205] Figure 6 A neighborhood diagram provided for an exemplary embodiment of the present application;

[0206] Figure 7 A schematic structural diagram of a data processing device provided for an exemplary embodiment of the present application;

[0207] Figure 8 A schematic structural diagram of another data processing device provided by an exemplary embodiment of the present application;

[0208] Figure 9 A schematic structural diagram of another data processing device provided by an exemplary embodiment of the present application;

[0209] Figure 10 A schematic structural diagram of another data processing device provided by an exemplary embodiment of the present application;

[0210] Figure 11 A structural diagram of a data processing device is provided for an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0211] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0212] To make the purpose, technical solutions, and advantages of this application more clear, the technical solutions of this application will be clearly and completely described below in conjunction with the specific embodiments of this application and the corresponding drawings. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0213] To facilitate understanding, the technical terms involved in the embodiments of this application are first explained.

[0214] Elastic Compute Service (ECS): is a secure, reliable, elastic and scalable cloud computing service.

[0215] STD stands for Season-Trend Decomposition, which decomposes a time series into a trend term (i.e., trend data), a seasonal term (i.e., cyclical data), and a residual term (i.e., residual data). This problem generally assumes that the trend term is slowly changing, the seasonal term is cyclical, and the residual term is white noise.

[0216] L1-norm-based optimization: The L1-norm is the sum of the absolute values of each element in a vector. Optimization problems with the L1-norm as the optimization objective generally have no closed-form solution and are computationally complex.

[0217] For ease of understanding, the following Figure 1 , describes the application scenarios involved in the embodiments of this application.

[0218] Figure 1 This is a schematic diagram of a scenario provided by an exemplary embodiment of this application. Figure 1 , performing STD processing on periodic time series data can obtain trend data, periodic data and residual data.

[0219] Currently, STD processing can be performed on periodic time series data using the RobustSTL algorithm or the OneShotSTL algorithm. Both the RobustSTL algorithm and the OneShotSTL algorithm include an optimization step based on the L1 norm. The inventors have discovered that the L1 norm-based optimization step is relatively complex. When performing STD processing on periodic time series data using the RobustSTL algorithm or the OneShotSTL algorithm, the L1 norm-based optimization step takes up a large amount of time, resulting in low data processing efficiency.

[0220] In view of this, an embodiment of the present application provides a data processing method, in which data can be decomposed through multiple simple algorithms to obtain corresponding trend data and / or periodic data without using an optimization process based on the L1 norm, thereby improving data processing efficiency.

[0221] The technical solutions shown in this application are described in detail below through specific embodiments. It should be noted that the following embodiments can exist independently or in combination with each other, and the same or similar contents will not be repeated in different embodiments.

[0222] Figure 2 This is a flow chart of a data processing method provided by an exemplary embodiment of the present application. Figure 2 , methods may include:

[0223] S201. Obtain first indicator data corresponding to a monitoring indicator.

[0224] The execution subject of this embodiment may be a data processing device, or a data processing apparatus provided within the data processing device. The data processing apparatus may be implemented via software or a combination of software and hardware. For ease of understanding, the following description will be based on an example in which the execution subject is a data processing device.

[0225] The monitoring indicator may be an operating parameter or a service parameter of the monitored device. For example, the device may be a server or a computer, and the monitoring indicator may be a central processing unit (CPU) occupancy rate or a memory occupancy rate.

[0226] The indicator data corresponding to the monitoring indicator can be the parameter values corresponding to the parameters of the monitored device (including operating parameters and business parameters). For example, if the monitoring indicator is CPU utilization, the indicator parameter corresponding to the monitoring indicator can be 45%. In other words, the indicator data corresponding to the monitoring indicator can be the data generated by the operation of the device.

[0227] The first indicator data corresponding to the monitoring indicator may be the maximum indicator data at the acquisition moment. In other words, the first indicator data corresponding to the monitoring indicator may be the most recently acquired data generated by the operation of the device.

[0228] Optionally, the first indicator data may be operation data or business data generated in a cloud computing scenario. For example, the first indicator data may be operation data of the ECS instance or business data of the ECS instance obtained by monitoring the operation of the ECS instance.

[0229] The data processing device may monitor the monitoring indicator to obtain the first indicator data; or the data processing device may receive the first indicator data sent by another device, for example, the other device may be a monitoring device.

[0230] For example, in order to analyze the operating stability of the ECS instance, the data processing device may monitor the operating status of the ECS instance and obtain data generated by the operation of the ECS instance; or, the data processing device may receive data generated by the operation of the ECS instance sent by the monitoring device.

[0231] S202: Determine multiple data processing algorithms.

[0232] The data processing algorithm may be an algorithm for performing STD processing on the first indicator data. For example, the data processing algorithm may be an algorithm for determining whether the indicator data has a jump, an algorithm for calculating a trend value corresponding to the indicator data, an algorithm for calculating a period value corresponding to the indicator data, etc.

[0233] In this embodiment, the algorithm complexity of the data processing algorithm is less than or equal to a preset complexity. The preset complexity may be the complexity of an optimization process based on the L1 norm. In other words, the plurality of data processing algorithms may be algorithms with a complexity less than that of an optimization process based on the L1 norm.

[0234] In this embodiment, the storage space of the data processing device may store multiple data processing algorithms. After the data processing device obtains the first indicator data, it may determine multiple data processing algorithms to perform STD processing on the first indicator data.

[0235] S203: Determine first historical indicator data corresponding to the first indicator data.

[0236] The first historical indicator data may be indicator data acquired at a time earlier than the time at which the first indicator data was acquired. In other words, the first historical indicator data may be indicator data corresponding to the monitoring indicator acquired before the first indicator data was acquired.

[0237] In this embodiment, the data processing device can obtain and store the indicator data corresponding to the monitoring indicator. In a specific implementation, the data processing device can determine the indicator data corresponding to the monitoring indicator that is obtained at a time later than the time when the first indicator data is obtained as the first historical indicator data corresponding to the first indicator data.

[0238] S204: Process the first indicator data and the first historical indicator data using multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator.

[0239] In this embodiment, the first indicator data and the first historical indicator data may be subjected to STD processing using a plurality of data processing algorithms. For example, the first indicator data and the first historical indicator data may be processed using an algorithm that calculates a trend value corresponding to the indicator data to obtain trend data corresponding to the monitoring indicator; or the first indicator data and the first historical indicator data may be processed using an algorithm that calculates a period value corresponding to the indicator data to obtain periodic data corresponding to the monitoring indicator.

[0240] The trend data corresponding to the monitoring indicator may include a trend value of the first indicator data and a trend value of the first historical indicator data obtained by performing STD processing on the first indicator data and the first historical indicator data.

[0241] The periodic data corresponding to the monitoring indicator may include a periodic value of the first indicator data and a periodic value of the first historical indicator data obtained by performing STD processing on the first indicator data and the first historical indicator data.

[0242] The first indicator data and the first historical indicator data are processed by a data processing algorithm with less complexity, which reduces the complexity of data processing and saves data processing resources.

[0243] In this embodiment, a data processing device can obtain first indicator data corresponding to a monitoring indicator and determine multiple data processing algorithms; the first indicator data and first historical indicator data can be processed using the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator. In the above method, the complexity of the multiple data processing algorithms is relatively low, resulting in a lower complexity in decomposing the indicator data (including the first indicator data and the first historical indicator data) into trend data and / or periodic data corresponding to the monitoring indicator, thereby improving data processing efficiency.

[0244] In actual applications, indicator data corresponding to monitoring indicators can be generated in real time or periodically. During an initial period, when the amount of indicator data corresponding to a monitoring indicator is relatively small, the indicator data can be accumulated until the number of indicators reaches a preset value M, at which point the M indicator data are processed. This process is referred to as the initialization process. After the initialization process, each time new indicator data is generated, it can be processed in real time. This process is referred to as the online update process.

[0245] It should be noted that the preset value M can be the number of indicator data included in a sliding window of fixed length. In this embodiment, a sliding window of fixed length can be maintained, and the sliding window can include multiple historical indicator data. During specific implementation, all historical indicator data in the sliding window can be processed each time to obtain trend data and / or periodic data corresponding to the monitoring indicator.

[0246] Optionally, M=(K+1)T, where T may be the number of indicator data contained in one period of the periodic time series, K may be the number of historical monitoring periods, and K may be an integer greater than or equal to 1. It should be understood that M may be an integer multiple of T.

[0247] Below, taking K equal to 2 as an example, combined with Figure 3 M will be described.

[0248] Figure 3 This is a schematic diagram of the quantity of indicator data in a sliding window provided by an exemplary embodiment of this application. Figure 3 , the sliding window may include M indicator data. The M indicator data may be indicator data collected within three monitoring cycles, wherein T indicator data may be collected within each monitoring cycle. In other words, the M indicator data may be indicator data included in three cycles of the periodic time series.

[0249] For ease of understanding, the initialization process and the online update process are described in detail below.

[0250] First, the initialization process is described.

[0251] During the initialization process, if the first data volume is less than or equal to M, when M indicator data corresponding to the monitoring indicator are obtained, the M indicator data are processed through multiple data processing algorithms to obtain trend data and / or periodic data, and the M indicator data include the first indicator data and the first historical indicator data.

[0252] Specifically, if the number of historical indicator data included in the sliding window is less than M, monitoring data corresponding to the monitoring indicator may continue to be acquired until the number of historical indicator data included in the sliding window is M. The data processing device may process the M indicator data to obtain trend data and / or periodic data.

[0253] It should be noted that the specific method of processing M indicator data can be found in Figure 4 The embodiments are not described in detail here.

[0254] After completing the initialization process, if the data processing device obtains the first indicator data (ie, new indicator data), it may enter the online update process and perform STD processing on the first indicator data.

[0255] Next, the online update process is described.

[0256] During the online update process, the historical indicator data included in the sliding window may include the first indicator data and M-1 first historical indicator data. The data processing device may process the first indicator data and M-1 first historical indicator data to obtain trend data and / or periodic data.

[0257] Specifically, during the online update process, if the data processing device obtains the first indicator data, it can remove the indicator data with the earliest acquisition time in the sliding window, add the first indicator data to the sliding window, and perform STD processing on the first indicator data. It should be understood that before the STD processing is performed on the first indicator data, the STD processing of M-1 first historical indicator data has been completed. It should be noted that the specific method of processing the first indicator data and M-1 first historical indicator data can be found in Figure 4 The embodiments are not described in detail here.

[0258] In this embodiment, through the initialization process and the online update process, each indicator data corresponding to the monitoring indicator can be subjected to STD processing to obtain trend data and / or periodic data corresponding to the monitoring indicator. The trend data includes trend values of multiple historical indicator data corresponding to the monitoring indicator, and the periodic data includes period values of multiple historical indicator data; for any historical indicator data, the sum of the trend value and period value of the historical indicator data is determined as the to-be-processed value corresponding to the historical indicator data; the difference between the historical indicator data and the to-be-processed value is determined as the residual value corresponding to the historical indicator data; and the residual data corresponding to the monitoring indicator is determined to include the residual value corresponding to each historical indicator data.

[0259] Below, in the above Figure 2 Based on the embodiment shown, combined Figure 4 , the data processing method provided in the embodiments of the present application is described in detail.

[0260] Figure 4 A flowchart of another data processing method provided by an exemplary embodiment of the present application. Figure 4 , methods may include:

[0261] S401. Obtain first indicator data corresponding to a monitoring indicator.

[0262] S402: Determine multiple data processing algorithms.

[0263] S403: Determine first historical indicator data corresponding to the first indicator data.

[0264] It should be noted that the specific implementation of S401-S403 can be found in S201-S203 and will not be repeated here.

[0265] S404: Determine a first data volume of historical indicator data corresponding to the monitoring indicator.

[0266] If the first data is less than or equal to M, execute S405;

[0267] If the first data is greater than M, execute S409.

[0268] It should be noted that S405-S408 correspond to the above-mentioned initialization process, and S409-S412 correspond to the above-mentioned online update process.

[0269] S405: When M indicator data corresponding to the monitoring indicator are obtained, at least one second indicator data is determined in the M indicator data.

[0270] The second indicator data is the indicator data that undergoes a jump among the M indicator data.

[0271] In this embodiment, when at least one second indicator data is determined among M indicator data, multiple third indicator data can be determined among the M indicator data, where the third indicator data is the data in the middle part of the M indicator data; for any third indicator data, a jump variable of the third indicator data is determined, where the jump variable is used to indicate the difference between the future indicator data and the historical indicator data of the third indicator data, and the generation time of the future indicator data is after the generation time of the third indicator data; based on the jump variable of each third indicator data, selected indicator data is determined among multiple third indicator data, where the jump variable of the selected indicator data is the local extreme value point among the jump variables of the multiple third indicator data; based on the M indicator data, a jump variable threshold corresponding to each selected indicator data is determined; indicator data whose absolute value of the jump variable in the selected indicator data is greater than the corresponding jump variable threshold is determined as at least one second indicator data.

[0272] In this embodiment, the M indicator data can be divided into three data segments according to the order of collection time, and the number of indicator data included in each data segment is the same. Then the data in the middle part can be the indicator data included in the second data segment. Figure 3 For example, the third indicator data may be T indicator data in the middle of the sliding window.

[0273] In this embodiment, the jump value of each third indicator data can be determined by the following formula 1:

[0274]

[0275] Among them, d i It can be the jump variable of the i-th indicator data among the M indicator data;

[0276] y i It can be the i-th indicator data among the M indicator data;

[0277] T may be the number of indicator data collected in a single monitoring cycle.

[0278] It should be noted that, through the above formula 1, the jump amount of each third indicator data can be determined to obtain the jump amounts of multiple third indicator data.

[0279] In this embodiment, the jump amount of the candidate indicator data may be a local extreme value point in the jump amount of the plurality of third indicator data. In other words, the third indicator data corresponding to the local extreme value point in the jump amount of the plurality of third indicator data may be the candidate indicator data.

[0280] Specifically, the jump amount of the selected indicator data may be a local maximum point or a local minimum point among the jump amounts of the plurality of third indicator data. For any local maximum point, the local maximum point may be greater than the jump amounts of the two adjacent third indicator data; for any local minimum point, the local minimum point may be less than the jump amounts of the two adjacent third indicator data.

[0281] In this embodiment, for any one of the candidate indicator data, if the jump variable of the candidate indicator data satisfies the following formula 2, the candidate indicator data can be determined as the second indicator data.

[0282] |d i |>n×max{std(y i ,…y i+T-1 ),std(y i-T ,…y i-1 )} (Formula 2)

[0283] Wherein, n can be a parameter of N-Sigma detection. Optionally, n can be 6.

[0284] That is to say, for any jump variable d of the selected indicator data i , d i The jump threshold can be n×max{std(y i ,…y i+T-1 ),std(y i-T ,…y i-1 )}.

[0285] The algorithm for determining the jump value of the third indicator data in S405 and the algorithm for determining the second indicator data may both be data processing algorithms.

[0286] S406. Segment the M indicator data according to at least one second indicator data to obtain multiple indicator data sequences.

[0287] In this embodiment, the M indicator data may be indicator data collected at M consecutive moments. The M indicator data may include at least one second indicator data. The at least one second indicator data may divide the M consecutive indicator data into multiple indicator data sequences.

[0288] Next, combine Figure 5 , explaining multiple indicator data series.

[0289] Figure 5 This is a schematic diagram of an indicator data sequence provided by an exemplary embodiment of this application. Figure 5 As shown, the M indicator data may include two second indicator data. The two second indicator data may be second indicator data 1 and second indicator data 2. The second indicator data 1 and the second indicator data 2 may be used to segment the M indicator data to obtain indicator data sequence 1, indicator data sequence 2, and indicator data sequence 3.

[0290] S407: Determine the trend value of each indicator data in each indicator data sequence.

[0291] In this embodiment, for any indicator data sequence, if the number of indicator data included in the indicator data sequence is different, the method for determining the trend value of each indicator data in the indicator data sequence is also different. The method for determining the trend value of each indicator data in the indicator data sequence includes at least the following two cases:

[0292] Case 1: For any indicator data sequence, if the number of indicator data included in the indicator data sequence is less than or equal to a first threshold T, the average value of each indicator data in the indicator data sequence is determined as the trend value of each indicator data in the indicator data sequence. Where T can be the number of indicator data collected in a single monitoring cycle.

[0293] In this case, the number of indicator data included in the indicator data sequence is less than or equal to the number of indicator data collected in a single monitoring cycle. The data processing device can determine the average value of each indicator data in the indicator data sequence as the trend value of each indicator data in the indicator data sequence.

[0294] For example, assuming that the indicator data sequence 2 includes m indicator data, where m is an integer greater than 1 and less than T, the trend value of each indicator data in the indicator data sequence 2 may be the average value of the m indicator data included in the indicator data sequence 2.

[0295] Case 2: For any indicator data sequence, if the number of indicator data included in the indicator data sequence is greater than a first threshold, a first subsequence and a second subsequence are determined in the indicator data sequence, and based on each indicator data in the indicator sequence, the trend value of each indicator data in the first subsequence is determined, and the trend value of each indicator data in the second subsequence is determined as the average value of each indicator data in the second subsequence; wherein, the second subsequence includes the last T-1 indicator data in the indicator data sequence, the first subsequence includes the other indicator data in the indicator data sequence except the second subsequence, and T is an integer greater than 1.

[0296] Specifically, for any indicator data y in the first subsequence 1n , the indicator data y 1n The trend value is denoted as τ 1n , and can determine τ 1n =y 1n +y 1n+1 +…+y 1n+T-1 .

[0297] For any indicator data y in the second subsequence 2n , the indicator data y 2n The trend value is denoted as τ 2n , and can determine τ 2n Equal to the average value of each indicator data in the second subsequence.

[0298] In this case, if the maximum indicator data at the time of acquisition in the indicator data sequence is y p , then the indicator data y p-T+1 Determine the subsequence segmentation index data. Among them, p can be y p The acquisition time, p-T+1 can be y p-T+1 The index data sequence with an acquisition time less than or equal to p-T+1 can form a first subsequence. The index data sequence with an acquisition time greater than p-T+1 can form a second subsequence.

[0299] In this embodiment, the trend value of each indicator data in each indicator data sequence can be determined according to the method of S407. In other words, the trend value of each indicator data in the M indicator data included in the sliding window during the initialization process can be determined according to the method of S407. The trend data can include the trend value of each indicator data in the M indicator data.

[0300] The algorithm for determining the trend value of each indicator data in each indicator data sequence in S407 may be a data processing algorithm.

[0301] S408. Determine the period value of each indicator data by using a non-local periodic filtering algorithm.

[0302] It should be noted that the method for determining the period value of any indicator data in any indicator data sequence is consistent. Below, taking any indicator data in any indicator data sequence as an example, the method for determining the period value of indicator data is described.

[0303] In this embodiment, multiple first reference indicator data corresponding to the indicator data can be determined among M indicator data; the weight value of each first reference indicator data can be determined; and the period value of the indicator data can be determined based on the weight value of each first reference indicator data, each first reference indicator data, and the trend value of each first reference indicator data.

[0304] Assume that the indicator data is y t , where t can be y t In this embodiment, the acquisition time can be t K neighborhoods are determined in the corresponding K historical monitoring cycles, where one neighborhood can be determined in each historical monitoring cycle. The center of the neighborhood can be t′=t-kT, and the width of the neighborhood on one side can be H. t The corresponding multiple first reference index data can be y′ i ,in,

[0305] y′ i =y i -τ i

[0306] i∈Ω, Ω={i|(t′=t-kT, i=t′±h)}, k=1, 2…, K, h=0, 1…, H

[0307] The weight value of each first reference indicator data can be in,

[0308]

[0309] in,

[0310]

[0311] δ may be the standard deviation of periodic data changes. It should be noted that δ may be set according to actual needs, and this embodiment does not limit this.

[0312] Therefore, the indicator data y t The period value can be s t ,in,

[0313]

[0314] It should be noted that the above method for determining the period value of the indicator data may be a non-local periodic filtering algorithm, and the non-local periodic filtering algorithm may be a data processing algorithm.

[0315] In this embodiment, the periodic value of each indicator data in the M indicator data may be determined in the manner of S408. The periodic data may include the periodic value of each indicator data in the M indicator data.

[0316] Below, taking K equal to 2 as an example, combined with Figure 6 , for y t The corresponding K neighborhoods are explained.

[0317] Figure 6 A neighborhood diagram is provided for an exemplary embodiment of the present application. Figure 6 As shown, assuming y t The acquisition time is t, then y t The center of the neighborhood in the first historical monitoring period can be tT,y t The center of the corresponding neighborhood in the second historical monitoring period may be t-2T. The width of each side of the neighborhood may be H. The starting time of the neighborhood in the first historical monitoring period may be tTH, the ending time of the neighborhood in the first historical monitoring period may be t-T+H, the starting time of the neighborhood in the second historical monitoring period may be t-2T-H, and the ending time of the neighborhood in the second historical monitoring period may be t-2T+H.

[0318] S409: Determine the predicted value corresponding to the first indicator data.

[0319] In this embodiment, multiple second reference indicator data corresponding to the first indicator data and the first trend value of the previous indicator data of the first indicator data can be determined among multiple historical indicator data of the indicator data; the predicted value is determined based on the first indicator data, the periodic value of each second reference indicator data, and the first trend value.

[0320] Assume that the first indicator data is y t1 , where t1 can be y t1 The acquisition time of the first indicator data y t1 The corresponding multiple second reference index data can be y i ,in,

[0321] i∈Ω, Ω={i|(t′=t-kT, i=t′±h)}, k=1, 2…, K, h=0, 1…, H

[0322] The period value of the second reference indicator data can be s i .

[0323] The previous indicator data of the first indicator data can be y t1-1 , the first trend value of the previous indicator data of the first indicator data can be τ t1-1 .

[0324] The predicted value corresponding to the first indicator data can be in,

[0325]

[0326] The algorithm for determining the predicted value corresponding to the first indicator data may be a data processing algorithm.

[0327] S410: Determine whether the first indicator data is outlier data based on the first indicator data and the predicted value.

[0328] If yes, execute S411;

[0329] If not, execute S412.

[0330] In this embodiment, M consecutive first historical indicator data may be determined as a set Y, if the first indicator data satisfies:

[0331]

[0332] Wherein, n can be a parameter of N-Sigma detection. Optionally, n can be 6.

[0333] It can be determined that the first indicator data is outlier data.

[0334] The algorithm for determining whether the first indicator data is outlier data may be a data processing algorithm.

[0335] In this embodiment, if the first indicator data is an outlier, the indicator data with the earliest acquisition time in set Y can be removed, and the predicted value corresponding to the first indicator data can be added to set Y to obtain an updated set Y′. If the first indicator data is not an outlier, the indicator data with the earliest acquisition time in set Y can be removed, and the first indicator data can be added to set Y to obtain an updated set Y′.

[0336] S411. Determine a first quantity of continuous outlier data, and determine trend data and / or periodic data based on the first quantity, first indicator data, and M-1 first historical indicator data.

[0337] In this embodiment, it can be determined whether the first number is greater than or equal to a preset L, where L is an integer greater than or equal to 1; if so, then among the M-1 first historical indicator data, the L-1 second historical indicator data before the first indicator data are determined, and based on the first indicator data and the L-1 second historical indicator data, the target trend value is determined, and the trend values of the first indicator data and the L-1 second historical indicator data are updated to the target trend value, and the period value of the first indicator data is determined by a non-local periodic filtering algorithm; if not, the predicted value and the average value of the M-1 first historical indicator data are determined as the trend value of the first indicator data, and the period value of the first indicator data is determined by a non-local periodic filtering algorithm; wherein, the trend data includes the trend value of the first indicator data, and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data, and the period value of each first historical indicator data.

[0338] Specifically, if the first number is greater than or equal to L, then:

[0339] The target trend value can be in,

[0340]

[0341] After the target trend value is determined, the trend value of the first indicator data and L-1 second historical indicator data may be determined as the target trend value.

[0342] The period value of the first indicator data can be s t1 ,in,

[0343]

[0344] i∈Ψ, Ψ={i|(t1'=t1-kT, i=t1′±h)}, k=1, 2..., K, h=0, 1..., H

[0345] y′ i =y i -τ i

[0346]

[0347]

[0348] If the first number is less than L, then:

[0349] The trend value of the first indicator data can be the average value of the predicted value and M-1 first historical indicator data. That is, the trend value τ1 of the first indicator data can be,

[0350] τ1=mean(Y′)

[0351] The set may include the predicted value and M-1 first historical indicator data.

[0352] The method for determining the period value of the first indicator data can refer to the case where the first number is greater than or equal to L, and will not be repeated here.

[0353] In this embodiment, the trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data. For the method of determining the trend value of each first historical indicator data and the period value of each first historical indicator data, refer to the method of determining the trend value of the first indicator data and the period value of the first historical indicator data, or refer to S405-S408.

[0354] The above-mentioned algorithm for determining the target trend value and the algorithm for determining the period value of the first indicator data may all be data processing algorithms.

[0355] In this embodiment, after determining the target trend value based on the first indicator data and L-1 second historical indicator data, the first indicator data and L-1 second historical indicator data can also be updated to the target trend value; and, for any other historical indicator data in the M-1 historical indicator data except the L-1 second historical indicator data, the update increment is determined based on the target trend data and the trend value of the Lth historical indicator data before the first indicator data, and the historical indicator data is updated to the sum of the historical indicator data and the update increment.

[0356] The trend value of the Lth historical indicator data before the first indicator data can be τ t1-L .

[0357] The update increment can be the difference between the target trend data and the trend value of the Lth historical indicator data before the first indicator data.

[0358] Specifically, the first indicator data and L-1 second historical indicator data in the set Y′ can be updated as Furthermore, any other historical indicator data in the set Y′ except the first indicator data and L-1 second historical indicator data can be updated to the historical indicator data and The harmony.

[0359] It should be noted that after the first indicator data and L-1 second historical indicator data are updated to the target trend value, the period value of the first indicator data and L-1 second historical indicator data can be updated through the non-local periodic filtering algorithm, and the residual value of the first indicator data and L-1 second historical indicator data can also be updated according to the updated trend value and period value.

[0360] In this embodiment, if the first number of continuous outlier data is greater than or equal to L, it can be determined that a trend jump has occurred in the indicator data corresponding to the monitoring indicator, and the trend values of the first indicator data and L-1 second historical indicator data can be updated according to the target trend value to reduce the impact of the trend jump on the STD processing, so that the STD processing effect of the indicator data corresponding to the monitoring indicator is better.

[0361] S412: Determine an average value of the first indicator data and M-1 first historical indicator data as a trend value of the first indicator data, and determine a period value of the first indicator data by using a non-local periodic filtering algorithm.

[0362] Specifically, the trend value τ1 of the first indicator data can be,

[0363] τ1=mean(Y′)

[0364] The set may include the first indicator data and M-1 first historical indicator data.

[0365] It should be noted that the method for determining the period value of the first indicator data can be found in S411 and will not be described in detail here.

[0366] In this embodiment, the trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data. For the method of determining the trend value of each first historical indicator data and the period value of each first historical indicator data, refer to the method of determining the trend value of the first indicator data and the period value of the first historical indicator data, or refer to S405-S408.

[0367] The algorithm for determining the periodicity of the first indicator data and the algorithm for determining the trend value of the first indicator data may both be data processing algorithms.

[0368] In this embodiment, the data processing device can obtain the first indicator data corresponding to the monitoring indicator and determine multiple data processing algorithms; can determine the first historical indicator data corresponding to the first indicator data; can determine the first data volume of the historical indicator data corresponding to the monitoring indicator; if the first data is less than or equal to M, when the M indicator data corresponding to the monitoring indicator are obtained, at least one second indicator data can be determined in the M indicator data; the M indicator data can be segmented and processed according to the at least one second indicator data to obtain multiple indicator data sequences; the trend value of each indicator data in each indicator data sequence can be determined; and the period value of each indicator data can be determined by a non-local periodic filtering algorithm. If the first number is less than M, the predicted value corresponding to the first indicator data can be determined, and whether the first indicator data is outlier data can be judged based on the first indicator data and the predicted value; if the first indicator data is outlier data, the first number of continuous outlier data can be determined, and trend data and / or periodic data can be determined based on the first number, the first indicator data and M-1 first historical indicator data; if the first indicator data is non-outlier data, the average value of the first indicator data and M-1 first historical indicator data can be determined as the trend value of the first indicator data, and the period value of the first indicator data can be determined by a non-local periodic filtering algorithm. In the above method, the complexity of multiple data processing algorithms is low, so that the complexity of decomposing the indicator data (including the first indicator data and the first historical indicator data) into the trend data and / or periodic data corresponding to the monitoring indicator is low, thereby improving data processing efficiency.

[0369] Figure 7 This is a schematic diagram of the structure of a data processing device provided by an exemplary embodiment of the present application. Figure 7 The data processing device 10 includes: an acquisition module 11, a determination module 12 and a processing module 13, wherein:

[0370] The acquisition module 11 is used to obtain first indicator data corresponding to the monitoring indicator;

[0371] The determining module 12 is used to determine a plurality of data processing algorithms, wherein the algorithm complexity of the data processing algorithms is less than or equal to a preset complexity, and the plurality of data processing algorithms are used to perform data decomposition processing on the data;

[0372] The determining module 12 is further configured to determine first historical indicator data corresponding to the first indicator data;

[0373] The processing module 13 is configured to process the first indicator data and the first historical indicator data using the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator.

[0374] The data processing device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be repeated here.

[0375] In a possible implementation, the processing module 13 is specifically configured to:

[0376] Determining a first data volume of historical indicator data corresponding to the monitoring indicator;

[0377] If the first data volume is less than or equal to M, then when M indicator data corresponding to the monitoring indicator are obtained, the M indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data, and the M indicator data include the first indicator data and the first historical indicator data;

[0378] If the first data volume is greater than M, the first indicator data and M-1 first historical indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data.

[0379] In a possible implementation, the processing module 13 is specifically configured to:

[0380] Determining at least one second indicator data from the M indicator data, where the second indicator data is indicator data that undergoes a jump among the M indicator data;

[0381] According to the at least one second indicator data, the M indicator data are segmented to obtain a plurality of indicator data sequences;

[0382] Determine the trend value of each indicator data in each indicator data series, and determine the period value of each indicator data by using a non-local periodic filtering algorithm;

[0383] Determine that the trend data includes the trend value of each indicator data in the M indicator data, and determine that the periodic data includes the period value of each indicator data in the M indicator data; wherein the multiple data processing algorithms include the algorithms used in the above steps.

[0384] In a possible implementation, the processing module 13 is specifically configured to:

[0385] Determining a plurality of third indicator data from the M indicator data, wherein the third indicator data is data in a middle portion of the M indicator data;

[0386] For any third indicator data, determining a jump value of the third indicator data, where the jump value is used to indicate a difference between future indicator data and historical indicator data of the third indicator data, where a generation time of the future indicator data is after a generation time of the third indicator data;

[0387] Determine, according to the jump amount of the third indicator data, candidate indicator data from the plurality of third indicator data, wherein the jump amount of the candidate indicator data is a local extreme value point in the jump amount of the plurality of third indicator data;

[0388] For any one candidate indicator data, determine the jump value threshold corresponding to the candidate indicator data according to the M indicator data;

[0389] Indicator data whose absolute value of the jump amount is greater than the corresponding jump amount threshold in the candidate indicator data is determined as the at least one second indicator data.

[0390] In a possible implementation, for any indicator data sequence, the processing module 13 is specifically configured to:

[0391] If the number of indicator data included in the indicator data sequence is less than or equal to a first threshold value T, then determining the average value of each indicator data in the indicator data sequence as the trend value of each indicator data in the indicator data sequence;

[0392] If the number of indicator data included in the indicator data sequence is greater than the first threshold, determining a first subsequence and a second subsequence in the indicator data sequence, and determining a trend value of each indicator data in the first subsequence based on each indicator data in the indicator data sequence, and determining the trend value of each indicator data in the second subsequence as an average value of each indicator data in the first subsequence;

[0393] The second subsequence includes the last T-1 indicator data in the indicator data sequence, the first subsequence includes other indicator data in the indicator data sequence except the second subsequence, and T is an integer greater than 1.

[0394] In a possible implementation, for any indicator data in any indicator data sequence, the processing module 13 is specifically configured to:

[0395] Determining, among the M indicator data, a plurality of first reference indicator data corresponding to the indicator data;

[0396] Determining a weight value of each first reference indicator data;

[0397] The period value of the indicator data is determined according to the weight value of each first reference indicator data, each first reference indicator data, and the trend value of each first reference indicator data.

[0398] In a possible implementation, the processing module 13 is specifically configured to:

[0399] Determining a predicted value corresponding to the first indicator data, and judging whether the first indicator data is outlier data based on the first indicator data and the predicted value;

[0400] If yes, determining a first number of continuous outlier data, and determining the trend data and / or the periodic data based on the first number, the first indicator data, and M-1 first historical indicator data;

[0401] If not, the average value of the first indicator data and the M-1 first historical indicator data is determined as the trend value of the first indicator data, and the period value of the first indicator data is determined by the non-local periodic filtering algorithm, wherein the trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0402] In a possible implementation, the processing module 13 is specifically configured to:

[0403] Determining whether the first number is greater than or equal to a preset L, where L is an integer greater than or equal to 1;

[0404] If so, then determine the L-1 second historical indicator data before the first indicator data among the M-1 first historical indicator data, and determine a target trend value based on the first indicator data and the L-1 second historical indicator data, and update the trend values of the first indicator data and the L-1 second historical indicator data to the target trend value, and determine the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0405] If not, determining the average value of the predicted value and the M-1 first historical indicator data as the trend value of the first indicator data, and determining the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0406] The trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0407] The data processing device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be repeated here.

[0408] Figure 8 A schematic diagram of the structure of another data processing device provided by an exemplary embodiment of the present application. Figure 7 On the basis of, the data processing device 10 may further include an update module 14, wherein,

[0409] The updating module 14 is used to update the first indicator data and the L-1 second historical indicator data to the target trend value;

[0410] For any other historical indicator data among the M-1 historical indicator data except the L-1 second historical indicator data, the determining module 12 is further configured to determine an update increment based on the target trend data and a trend value of the Lth historical indicator data before the first indicator data;

[0411] The updating module 14 is further configured to update the historical indicator data to the sum of the historical indicator data and the update increment.

[0412] In a possible implementation, the processing module 13 is specifically configured to:

[0413] Determining, from the plurality of historical indicator data of the indicator data, a plurality of second reference indicator data corresponding to the first indicator data and a first trend value of indicator data preceding the first indicator data;

[0414] The predicted value is determined according to the first indicator data, the period value of each second reference indicator data, and the first trend value.

[0415] In a possible implementation, the trend data includes trend values of a plurality of historical indicator data corresponding to the monitoring indicator, and the periodic data includes period values of the plurality of historical indicator data; the determining module 12 is further configured to:

[0416] For any historical indicator data, the sum of the trend value and the period value of the historical indicator data is determined as the to-be-processed value corresponding to the historical indicator data;

[0417] Determine the difference between the historical indicator data and the value to be processed as the residual value corresponding to the historical indicator data;

[0418] Determine that the residual data corresponding to the monitoring indicator includes the residual value corresponding to each historical indicator data.

[0419] The data processing device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be repeated here.

[0420] Figure 9 This is a structural diagram of another data processing device provided by an exemplary embodiment of the present application. Figure 9 The data processing device 20 includes: an acquisition module 21, a determination module 22 and a processing module 23, wherein:

[0421] The acquisition module 21 is used to acquire first indicator data corresponding to the monitoring indicator, where the first indicator data is data generated by the operation of the device;

[0422] The determining module 22 is used to determine a plurality of data processing algorithms, wherein the algorithm complexity of the data processing algorithms is less than or equal to a preset complexity, and the plurality of data processing algorithms are used to perform data decomposition processing on the data;

[0423] The determining module 22 is further configured to determine first historical indicator data corresponding to the first indicator data;

[0424] The processing module 23 is configured to process the first indicator data and the first historical indicator data using the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator.

[0425] The data processing device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be repeated here.

[0426] In a possible implementation, the processing module 23 is specifically configured to:

[0427] Determining a first data volume of historical indicator data corresponding to the monitoring indicator;

[0428] If the first data volume is less than or equal to M, then when M indicator data corresponding to the monitoring indicator are obtained, the M indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data, and the M indicator data include the first indicator data and the first historical indicator data;

[0429] If the first data volume is greater than M, the first indicator data and M-1 first historical indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data.

[0430] In a possible implementation, the processing module 23 is specifically configured to:

[0431] Determining at least one second indicator data from the M indicator data, where the second indicator data is indicator data that undergoes a jump among the M indicator data;

[0432] According to the at least one second indicator data, the M indicator data are segmented to obtain a plurality of indicator data sequences;

[0433] Determine the trend value of each indicator data in each indicator data series, and determine the period value of each indicator data by using a non-local periodic filtering algorithm;

[0434] Determine that the trend data includes the trend value of each indicator data in the M indicator data, and determine that the periodic data includes the period value of each indicator data in the M indicator data; wherein the multiple data processing algorithms include the algorithms used in the above steps.

[0435] In a possible implementation, the processing module 23 is specifically configured to:

[0436] Determining a plurality of third indicator data from the M indicator data, wherein the third indicator data is data in a middle portion of the M indicator data;

[0437] For any third indicator data, determining a jump value of the third indicator data, where the jump value is used to indicate a difference between future indicator data and historical indicator data of the third indicator data, where a generation time of the future indicator data is after a generation time of the third indicator data;

[0438] Determine, according to the jump amount of the third indicator data, candidate indicator data from the plurality of third indicator data, wherein the jump amount of the candidate indicator data is a local extreme value point in the jump amount of the plurality of third indicator data;

[0439] For any one candidate indicator data, determine the jump value threshold corresponding to the candidate indicator data according to the M indicator data;

[0440] Indicator data whose absolute value of the jump amount is greater than the corresponding jump amount threshold in the candidate indicator data is determined as the at least one second indicator data.

[0441] In a possible implementation, for any indicator data sequence, the processing module 23 is specifically configured to:

[0442] If the number of indicator data included in the indicator data sequence is less than or equal to a first threshold value T, then determining the average value of each indicator data in the indicator data sequence as the trend value of each indicator data in the indicator data sequence;

[0443] If the number of indicator data included in the indicator data sequence is greater than the first threshold, determining a first subsequence and a second subsequence in the indicator data sequence, and determining a trend value of each indicator data in the first subsequence based on each indicator data in the indicator data sequence, and determining the trend value of each indicator data in the second subsequence as an average value of each indicator data in the first subsequence;

[0444] The second subsequence includes the last T-1 indicator data in the indicator data sequence, the first subsequence includes other indicator data in the indicator data sequence except the second subsequence, and T is an integer greater than 1.

[0445] In a possible implementation, for any indicator data in any indicator data sequence, the processing module 23 is specifically configured to:

[0446] Determining, among the M indicator data, a plurality of first reference indicator data corresponding to the indicator data;

[0447] Determining a weight value of each first reference indicator data;

[0448] The period value of the indicator data is determined according to the weight value of each first reference indicator data, each first reference indicator data, and the trend value of each first reference indicator data.

[0449] In a possible implementation, the processing module 23 is specifically configured to:

[0450] Determining a predicted value corresponding to the first indicator data, and judging whether the first indicator data is outlier data based on the first indicator data and the predicted value;

[0451] If yes, determining a first number of continuous outlier data, and determining the trend data and / or the periodic data based on the first number, the first indicator data, and M-1 first historical indicator data;

[0452] If not, the average value of the first indicator data and the M-1 first historical indicator data is determined as the trend value of the first indicator data, and the period value of the first indicator data is determined by the non-local periodic filtering algorithm, wherein the trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0453] In a possible implementation, the processing module 23 is specifically configured to:

[0454] Determining whether the first number is greater than or equal to a preset L, where L is an integer greater than or equal to 1;

[0455] If so, then determine the L-1 second historical indicator data before the first indicator data among the M-1 first historical indicator data, and determine a target trend value based on the first indicator data and the L-1 second historical indicator data, and update the trend values of the first indicator data and the L-1 second historical indicator data to the target trend value, and determine the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0456] If not, determining the average value of the predicted value and the M-1 first historical indicator data as the trend value of the first indicator data, and determining the period value of the first indicator data by using the non-local periodic filtering algorithm;

[0457] The trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

[0458] The data processing device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be repeated here.

[0459] Figure 10 A structural diagram of another data processing device provided by an exemplary embodiment of the present application. Figure 9 On the basis of, the data processing device 20 may further include an updating module 24, wherein,

[0460] The updating module 24 is used to update the first indicator data and the L-1 second historical indicator data to the target trend value;

[0461] For any other historical indicator data among the M-1 historical indicator data except the L-1 second historical indicator data, the determining module 22 is further configured to determine an update increment based on the target trend data and a trend value of the Lth historical indicator data before the first indicator data;

[0462] The updating module 24 is further configured to update the historical indicator data to the sum of the historical indicator data and the update increment.

[0463] In a possible implementation, the processing module 23 is specifically configured to:

[0464] Determining, from the plurality of historical indicator data of the indicator data, a plurality of second reference indicator data corresponding to the first indicator data and a first trend value of indicator data preceding the first indicator data;

[0465] The predicted value is determined according to the first indicator data, the period value of each second reference indicator data, and the first trend value.

[0466] In a possible implementation, the trend data includes trend values of a plurality of historical indicator data corresponding to the monitoring indicator, and the periodic data includes period values of the plurality of historical indicator data; the determining module 22 is further configured to:

[0467] For any historical indicator data, the sum of the trend value and the period value of the historical indicator data is determined as the to-be-processed value corresponding to the historical indicator data;

[0468] Determine the difference between the historical indicator data and the value to be processed as the residual value corresponding to the historical indicator data;

[0469] Determine that the residual data corresponding to the monitoring indicator includes the residual value corresponding to each historical indicator data.

[0470] The data processing device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be repeated here.

[0471] Figure 11 A schematic diagram of the structure of a data processing device is provided for an exemplary embodiment of the present application. Figure 11 The data processing device 30 may include a processor 31 and a memory 32. Exemplarily, the processor 31 and the memory 32 are interconnected via a bus 33.

[0472] The memory 32 stores computer-executable instructions;

[0473] The processor 31 executes the computer-executable instructions stored in the memory 32 , so that the processor 31 performs the method shown in the above method embodiment.

[0474] Figure 11 The data processing device shown in the figure may be a server, a computer or other electronic device.

[0475] The data processing device provided in the embodiment of the present application can execute the technical solution shown in the above method embodiment. Its implementation principles and beneficial effects are similar and will not be repeated here.

[0476] Accordingly, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the method described in the above method embodiment.

[0477] Accordingly, an embodiment of the present application may also provide a computer program product, including a computer program, which, when executed by a processor, may implement the method shown in the above method embodiment.

[0478] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0479] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0480] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0481] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0482] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0483] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0484] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0485] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0486] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A data processing method, characterized in that: include: Obtaining first indicator data corresponding to the monitoring indicator; Determining a plurality of data processing algorithms, wherein the algorithm complexity of the data processing algorithms is less than or equal to a preset complexity, and the plurality of data processing algorithms are used to perform data decomposition processing on the data; Determine first historical indicator data corresponding to the first indicator data, and process the first indicator data and the first historical indicator data through the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator.

2. A data processing method, characterized in that: include: Obtaining first indicator data corresponding to the monitoring indicator, where the first indicator data is data generated by the operation of the device; Determining a plurality of data processing algorithms, wherein the algorithm complexity of the data processing algorithms is less than or equal to a preset complexity, and the plurality of data processing algorithms are used to perform data decomposition processing on the data; Determine first historical indicator data corresponding to the first indicator data, and perform data decomposition processing on the first indicator data and the first historical indicator data through the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator.

3. The method according to claim 1 or 2, characterized in that Processing the first indicator data and the first historical indicator data by the multiple data processing algorithms to obtain trend data and / or periodic data corresponding to the monitoring indicator includes: Determining a first data volume of historical indicator data corresponding to the monitoring indicator; If the first data volume is less than or equal to M, then when M indicator data corresponding to the monitoring indicator are obtained, the M indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data, and the M indicator data include the first indicator data and the first historical indicator data; If the first data volume is greater than M, the first indicator data and M-1 first historical indicator data are processed by the multiple data processing algorithms to obtain the trend data and / or the periodic data.

4. The method according to claim 3, characterized in that Processing the M indicator data using the multiple data processing algorithms to obtain the trend data and / or the periodic data includes: Determining at least one second indicator data from the M indicator data, where the second indicator data is indicator data that undergoes a jump among the M indicator data; According to the at least one second indicator data, the M indicator data are segmented to obtain a plurality of indicator data sequences; Determine the trend value of each indicator data in each indicator data series, and determine the period value of each indicator data by using a non-local periodic filtering algorithm; Determine that the trend data includes the trend value of each indicator data in the M indicator data, and determine that the periodic data includes the period value of each indicator data in the M indicator data; wherein the multiple data processing algorithms include the algorithms used in the above steps.

5. The method according to claim 4, characterized in that Determining at least one second indicator data from the M indicator data includes: Determining a plurality of third indicator data from the M indicator data, wherein the third indicator data is data in a middle portion of the M indicator data; For any third indicator data, determining a jump value of the third indicator data, where the jump value is used to indicate a difference between future indicator data and historical indicator data of the third indicator data, where a generation time of the future indicator data is after a generation time of the third indicator data; Determine, according to the jump amount of the third indicator data, candidate indicator data from the plurality of third indicator data, wherein the jump amount of the candidate indicator data is a local extreme value point in the jump amount of the plurality of third indicator data; For any one candidate indicator data, determine the jump value threshold corresponding to the candidate indicator data according to the M indicator data; Indicator data whose absolute value of the jump amount is greater than the corresponding jump amount threshold in the candidate indicator data is determined as the at least one second indicator data.

6. The method according to claim 4 or 5, characterized in that For any indicator data sequence; determining the trend value of each indicator data in the indicator data sequence, including: If the number of indicator data included in the indicator data sequence is less than or equal to a first threshold value T, then determining the average value of each indicator data in the indicator data sequence as the trend value of each indicator data in the indicator data sequence; If the number of indicator data included in the indicator data sequence is greater than the first threshold, determining a first subsequence and a second subsequence in the indicator data sequence, and determining a trend value of each indicator data in the first subsequence based on each indicator data in the indicator data sequence, and determining the trend value of each indicator data in the second subsequence as an average value of each indicator data in the second subsequence; The second subsequence includes the last T-1 indicator data in the indicator data sequence, the first subsequence includes other indicator data in the indicator data sequence except the second subsequence, and T is an integer greater than 1.

7. The method according to claim 4 or 5, characterized in that For any indicator data in any indicator data sequence, determining the period value of the indicator data by using a non-local periodic filtering algorithm includes: Determining, among the M indicator data, a plurality of first reference indicator data corresponding to the indicator data; Determining a weight value of each first reference indicator data; The period value of the indicator data is determined according to the weight value of each first reference indicator data, each first reference indicator data, and the trend value of each first reference indicator data.

8. The method according to any one of claims 2 to 6, characterized in that: Processing the first indicator data and M-1 first historical indicator data by the multiple data processing algorithms to obtain the trend data and / or the periodic data includes: Determining a predicted value corresponding to the first indicator data, and judging whether the first indicator data is outlier data based on the first indicator data and the predicted value; If yes, determining a first number of continuous outlier data, and determining the trend data and / or the periodic data based on the first number, the first indicator data, and M-1 first historical indicator data; If not, the average value of the first indicator data and the M-1 first historical indicator data is determined as the trend value of the first indicator data, and the period value of the first indicator data is determined by the non-local periodic filtering algorithm, wherein the trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

9. The method according to claim 8, characterized in that Determining the trend data and / or the periodic data according to the first quantity, the first indicator data, and M-1 first historical indicator data includes: Determining whether the first number is greater than or equal to a preset L, where L is an integer greater than or equal to 1; If so, then determine the L-1 second historical indicator data before the first indicator data among the M-1 first historical indicator data, and determine a target trend value based on the first indicator data and the L-1 second historical indicator data, and update the trend values of the first indicator data and the L-1 second historical indicator data to the target trend value, and determine the period value of the first indicator data by using a non-local periodic filtering algorithm; If not, determining the average value of the predicted value and the M-1 first historical indicator data as the trend value of the first indicator data, and determining the period value of the first indicator data by using the non-local periodic filtering algorithm; The trend data includes the trend value of the first indicator data and the trend value of each first historical indicator data, and the periodic data includes the period value of the first indicator data and the period value of each first historical indicator data.

10. The method according to claim 9, characterized in that After determining the target trend value based on the first indicator data and the L-1 second historical indicator data, the method further includes: Updating the first indicator data and the L-1 second historical indicator data to the target trend value; and For any other historical indicator data among the M-1 historical indicator data except the L-1 second historical indicator data, the update increment is determined based on the target trend data and the trend value of the Lth historical indicator data before the first indicator data, and the historical indicator data is updated to the sum of the historical indicator data and the update increment.

11. The method according to any one of claims 8 to 10, characterized in that: Determining a predicted value corresponding to the first indicator data includes: Determining, from the plurality of historical indicator data of the indicator data, a plurality of second reference indicator data corresponding to the first indicator data and a first trend value of indicator data preceding the first indicator data; The predicted value is determined according to the first indicator data, the period value of each second reference indicator data, and the first trend value.

12. The method according to any one of claims 1 to 11, characterized in that The trend data includes trend values of a plurality of historical indicator data corresponding to the monitoring indicator, and the periodic data includes period values of the plurality of historical indicator data; the method further includes: For any historical indicator data, the sum of the trend value and the period value of the historical indicator data is determined as the to-be-processed value corresponding to the historical indicator data; Determine the difference between the historical indicator data and the value to be processed as the residual value corresponding to the historical indicator data; Determine that the residual data corresponding to the monitoring indicator includes the residual value corresponding to each historical indicator data.

13. A data processing device, characterized in that: include: at least one processor; as well as a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the data processing device to perform the method according to any one of claims 1 to 12.

14. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, and when a processor executes the computer-executable instructions, the method according to any one of claims 1 to 12 is implemented.

15. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 12 is implemented.

Images